In this letter to the editor,1 the authors raise a serious concern about the selection bias in the study by Sharara, et al,2 with the results published in Clinical Infectious Diseases.
In the study, 186 patients with pyelonephritis caused by extended-spectrum beta-lactamase (ESB)L-producing organisms were included and randomly assigned to receive either piperacillin-tazobactam (n=45) or carbapenem (n=141).2 The results showed no differences between the 2 groups with regard to clinical symptoms resolved by day 7 or 30-day mortality. Therefore, this study concluded that in the treatment of patients with pyelonephritis caused by ESBL-producing organisms, piperacillin-tazobactam may be a suitable alternative.
However, the letter to the editor raises concerns about this conclusion, claiming there was selection bias. In the study, patients with pyelonephritis received either piperacillin-tazobactam or carbapenem within 48 hours of their initial urine culture; however, of these patients, only patients who remained on the study drug (piperacillin-tazobactam or a carbapenem) for the duration of therapy were included. Patients who did not remain on the study drug after 72 hours and transitioned to an alternate arm of antibiotics (piperacillin-tazobactam to ertapenem) were excluded.
Because patients who do not clinically respond to an empiric antibiotic after 48 to 72 hours are commonly switched to other appropriate antibiotics, it is possible that patients who transitioned to another appropriate antibiotic due to poor clinical response to piperacillin-tazobactam were excluded from the study. This means that the remaining patients in the piperacillin-tazobactam group may be highly selective, which may have contributed to the clinical efficacy of piperacillin-tazobactam being comparable to carbapenem. To avoid this bias, the letter to the editor authors suggested that the study researchers include an intention-to-treat analysis of their data.
Overall, the letter to the editor concluded that, “the findings of Sharara’s study should be interpreted cautiously and further study is warranted to confirm the clinical efficacy of TZP in the treatment of ESBL-producing pyelonephritis.”
1. Wang JH and Lai CC. Is piperacillin-tazobactam really comparable to carbapenem for the treatment of pyelonephritis caused by ESBL-producing organisms? [published online February 5, 2020]. Clin Infect Dis. doi:10.1093/cid/ciaa115/5722397
2. Sharara SL, Amoah J, Pana ZD, Simner PJ, Cosgrove SE, Tamma PD. Is piperacillin-tazobactam really comparable to carbapenem for the treatment of pyelonephritis caused by ESBL-producing organisms? [published online December 20, 2019]. Clin Infect Dis. doi:10.1093/cid/ciz1205